1:11 p.m.
Hi,
the function reference says that after restricting the alphas in a VECM
you don't get $xtxinv, and I can see the reason why (I think).
However, this seems to work:
<hansl>
open denmark
vecm 4 1 LRM LRY IBO
restrict
a[1] = 0
b[1] = 1
end restrict --full
eval $coeff
eval $stderr
eval $xtxinv
</hansl>
Bug or feature?
thanks
sven
8:34 a.m.
Am 17.09.2020 um 19:11 schrieb Sven Schreiber:
Hi,
the function reference says that after restricting the alphas in a VECM
you don't get $xtxinv, and I can see the reason why (I think).
However, this seems to work:
<hansl>
open denmark
vecm 4 1 LRM LRY IBO
restrict
a[1] = 0
b[1] = 1
end restrict --full
eval $coeff
eval $stderr
eval $xtxinv
</hansl>
Bug or feature?
Ping - I don't think this has been clarified.
thanks
sven
1:31 p.m.
Am 18.11.2020 um 14:34 schrieb Sven Schreiber:
Am 17.09.2020 um 19:11 schrieb Sven Schreiber:
> the function reference says that after restricting the alphas in a VECM
> you don't get $xtxinv, and I can see the reason why (I think).
>
> However, this seems to work:
> ...
>
> eval $xtxinv
>
Ping - I don't think this has been clarified.
OK I checked this more concretely, and it seems $xtxinv in this
restricted alpha case indeed still refers to the _common_ regressors,
then excluding the EC term. That means the dimension of $xtxinv in the
unrestricted case is bigger than in the restricted case.
I think something needs fixing here, at least the documentation. But the
"easiest" thing would be to follow the documentation and throw an error
when this accessor is requested after a restricted VECM. Other opinions?
thanks
sven
1:36 p.m.
On Tue, 24 Nov 2020, Sven Schreiber wrote:
I think something needs fixing here, at least the documentation. But
the
"easiest" thing would be to follow the documentation and throw an error
when this accessor is requested after a restricted VECM. Other opinions?
I concur. This statistic doesn't make much sense in the context of VECM
estimation, and if you absolutely need to have it, a few hansl lines
should suffice.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
9:01 a.m.
Am 24.11.2020 um 19:36 schrieb Riccardo (Jack) Lucchetti:
On Tue, 24 Nov 2020, Sven Schreiber wrote:
> I think something needs fixing here, at least the documentation. But the
> "easiest" thing would be to follow the documentation and throw an error
> when this accessor is requested after a restricted VECM. Other opinions?
I concur. This statistic doesn't make much sense in the context of VECM
estimation, and if you absolutely need to have it, a few hansl lines
should suffice.
I checked again with a Dec 12th snapshot, this is still (erroneously, in
our view) returning the $xtxinv. Should I file a ticket?
thanks
sven
1422
days inactive
1522
days old
4 comments
2 participants
participants (2)
-
Riccardo (Jack) Lucchetti -
Sven Schreiber